OpenAI’s GPT-5.6 rollout is already a meaningful news story, but the access details matter more than the headline. If you are trying to figure out where GPT-5.6 can actually be used today, the short answer is that OpenAI has opened a limited preview through the API and Codex for approved organizations, while ChatGPT is not included during the preview. OpenAI’s preview note is the clearest source for that distinction.
What OpenAI says is available right now
OpenAI says GPT-5.6 is a limited preview, not a public launch. The company says Sol, Terra, and Luna are available through the OpenAI API and Codex to a small group of trusted partners and organizations. OpenAI also says there is no public application or waitlist, and that the preview is not open to individual consumers.
That means the access story is simple, even if it is not convenient: if your organization was not invited, you do not have self-serve access yet.
How to access GPT-5.6 in Codex and the API
According to OpenAI’s help documentation, access is scoped separately. Approval for the API does not automatically include Codex, and Codex approval does not automatically include the API. In other words, the organization, workspace, and product need to be approved for the specific surface you plan to use. OpenAI’s preview guide says participation is limited to selected organizations working with an account representative.
- If you have an OpenAI account representative, confirm whether your organization was invited.
- Check whether approval covers the API, Codex, or both.
- Use the approved API organization or Codex workspace exactly as provisioned.
For teams already inside the preview, the practical next step is not just model testing. It is deciding which workflows are worth moving first, because preview access is still constrained and may not cover every team or use case equally.
Why ChatGPT access is the confusing part
The phrase “GPT-5.6 in ChatGPT” is where many readers will get tripped up. OpenAI explicitly says GPT-5.6 is not available in ChatGPT during the preview. The company says it plans to expand availability as soon as possible, but it has not announced a general-availability date.
That also means you should not assume a ChatGPT plan alone will unlock it. OpenAI’s current guidance points to broader rollout later, not immediate consumer access now.
What businesses should do before broader rollout
For business teams, the useful move is to treat GPT-5.6 as a planning signal, not a procurement trigger. If your workflow depends on frontier models for coding, computer use, knowledge work, or cybersecurity-adjacent tasks, you should map the specific jobs you would actually hand to a model and decide whether they belong in ChatGPT, Codex, or the API.
- Identify one workflow that would benefit from stronger reasoning or automation.
- Check whether the workflow needs a chat interface, a coding environment, or an API integration.
- Write down the approval path and owner for each system you might use.
- Prepare a fallback plan if GPT-5.6 stays limited while you need to ship now.
The best response to a limited preview is usually not to wait. It is to line up the rollout so that when access expands, the organization already knows where the model belongs.
Bottom line: GPT-5.6 is real news, but the current access story is narrow. Approved organizations can use it in the API and Codex; ChatGPT is not part of the preview yet.
Sources
- Previewing GPT-5.6 Sol: a next-generation model — OpenAI
- A preview of GPT-5.6 Sol, Terra, and Luna — OpenAI Help Center
- Models — OpenAI API Docs
- Additional safety checks for biological and cybersecurity requests in API and Codex — OpenAI Help Center